European languages before Indo-European

Discussion of natural languages, or language in general.
User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

I am currently working on a dictionary of possible Aquan loanwords in western IE languages. There is still a lot of work to do - I am currently about halfway through Matasović's Etymological Dictionary of Proto-Celtic, and then comes Germanic - but some patterns begin to form which are broadly in agreement with my notion of a Para-Indo-European language branching off before ablaut. There are a few words which resemble IE words but are somewhat "off" phonologically or semantically, and may point at Urverwandtschaft, such as Aq *makwa 'son' vs. PIE *meh2ḱ- 'to raise, to grow'.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

Zju
Lebom
Lebom
Posts: 243
Joined: Tue May 08, 2012 11:10 am

Re: European languages before Indo-European

Post by Zju »

WeepingElf wrote:I am currently working on a dictionary of possible Aquan loanwords in western IE languages. There is still a lot of work to do - I am currently about halfway through Matasović's Etymological Dictionary of Proto-Celtic, and then comes Germanic - but some patterns begin to form which are broadly in agreement with my notion of a Para-Indo-European language branching off before ablaut. There are a few words which resemble IE words but are somewhat "off" phonologically or semantically, and may point at Urverwandtschaft, such as Aq *makwa 'son' vs. PIE *meh2ḱ- 'to raise, to grow'.
That sounds interesting. Do you plan to publish the results anywhere?

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

Zju wrote:
WeepingElf wrote:I am currently working on a dictionary of possible Aquan loanwords in western IE languages. There is still a lot of work to do - I am currently about halfway through Matasović's Etymological Dictionary of Proto-Celtic, and then comes Germanic - but some patterns begin to form which are broadly in agreement with my notion of a Para-Indo-European language branching off before ablaut. There are a few words which resemble IE words but are somewhat "off" phonologically or semantically, and may point at Urverwandtschaft, such as Aq *makwa 'son' vs. PIE *meh2ḱ- 'to raise, to grow'.
That sounds interesting. Do you plan to publish the results anywhere?
I'll put it on a new section of my web site, together with other Paleo-European stuff I am also working on.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

I am done with the Celtic etymological dictionary; it's now at 194 entries, including 18 candidates for Aquan-IE Urverwandtschaft. Next, I will go through Kroonen's Etymological Dictionary of Proto-Germanic, but now I need a break.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
mèþru
Smeric
Smeric
Posts: 1984
Joined: Thu Oct 29, 2015 6:44 am
Location: suburbs of Mrin
Contact:

Re: European languages before Indo-European

Post by mèþru »

What's the URL to your website?
ìtsanso, God In The Mountain, may our names inspire the deepest feelings of fear in urkos and all his ilk, for we have saved another man from his lies! I welcome back to the feast hall kal, who will never gamble again! May the eleven gods bless him!
kårroť

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

mèþru wrote:What's the URL to your website?
My web site is at www.joerg-rhiemeier.de, but the paleolinguistics section is under construction and not online yet.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

While the work on the web site and the substratum dictionary is under way, I shall share some thoughts with you (later to be rolled up into an article on the web site).

The Iberian Peninsula as a test case for pre-IE linguistic diversity

A few years ago, I posted an estimate of the linguistic diversity of pre-IE Europe, where I arrived at a figure of 18 to 30 stocks in all of Europe, with about 3 to 5 stocks on each of the three great Mediterranean peninsulas. Salmoneus, however, posted a much lower estimate according to which there were hardly a dozen stocks in all of Europe, and only one on each peninsula, and his arguments are not easily refuted.

Now we have a possible test case for this question: the Iberian Peninsula. This is the only place in western Europe where one of the old stocks survived in the form of Basque, and there are epigraphic remains of two further non-IE languages, Iberian and Tartessian (some say that Tartessian was Celtic and thus IE, but these claims have been rejected by most Celticists, and I am very skeptical about this as well; see below).

The question is: are these languages related, or not? If they are all related to each other, this would be evidence of a single stock in the peninsula. If they are all unrelated to each other, we would have at least three stocks. There could have been one or two more in the northwest of the peninsula, where no pre-IE language is attested. At least, these three languages do not form a close-knit group like Germanic or Slavic where the relationship is readily apparent to the non-specialist. They are at most more distantly related, forming a stock like IE or Uralic, but they may constitute two or three different stocks.

The idea that Basque and Iberian are related is old, but the 19th- to early 20th-century claims by people like Wilhelm von Humboldt or Hugo Schuchardt (who reconstructed two nicely matching nominal declension paradigms for Basque and Iberian - which later both turned out to be wrong) were premature, as virtually nothing was known about the history of Basque and the Iberian script was undeciphered. Only in the middle of the 20th century, with Koldo Mitxelena's reconstruction of Proto-Basque and Manuel Gómez-Moreno's decipherment of the Iberian script, meaningful comparisons became possible - and they yielded only little evidence in favour of a relationship. A few Iberian word elements (such as the toponym Iliberis, which looks much like Proto-Basque *ili berri 'new town') can be interpreted by means of Basque, but we don't know whether these interpretations are correct, and many others look nothing like Basque. And even if there are true resemblances, borrowing may be an alternative (something like *ili 'town' may be an Iberian loanword in Basque, though this is not equally likely with a basic adjective like *berri 'new').

Then there is Tartessian, a tough nut to crack. It looks nothing like Basque or Iberian. We don't even know whether it is IE or not. At first glance, John T. Koch's suggestion that Tartessian was a Celtic language seems to make sense as there is no shortage of Celtic names in the region. There are several towns with names in *-briga, a tribe named Celtici and the legendary king Arganthonios - all of which are as Celtic as they could be. But I have been informed that most Celticists remain skeptical, and that Koch's reading of the Tartessian inscriptions as Celtic is fraught with problems such as taking recourse to forms attested only in Goidelic. (I am not a Celticist and cannot judge whether this skepticism is justified or not.) Also, these inscriptions are mostly dated to the early 7th century BC, making them the oldest of all inscriptions found in the Iberian Peninsula, and it seems unlikely that Celtic spread that far that early. This makes it more likely that the Tartessian inscriptions represent a pre-Celtic, non-IE language, and that the Celts put an end to Tartessian literacy when they arrived around 500 BC.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
Salmoneus
Sanno
Sanno
Posts: 3197
Joined: Thu Jan 15, 2004 5:00 pm
Location: One of the dark places of the world

Re: European languages before Indo-European

Post by Salmoneus »

Nobody can know, of course.

That said, I'm afraid I'm not convinced.

So far as I can see, you need three premises:
- there was linguistic diversity in Iberia
- that linguistic diversity preceded the kurgan invasions
- in terms of diversity, Iberia was representative of Europe as a whole

All three premises, in my opinion, are highly questionable.

Regarding the three families and diversity in Iberia:
- it's not even clear that Basque was present in Iberia in prehistoric times; our earliest and most voluminous attestations are from France, not Spain, and it appears that in at least part of its later Iberian territory it was overlaid on a Celtic (or similar) substrate. While it's possible that it was present all along, we can't really assume that for sure, as I understand it.

- Iberian appears to be related to Basque. Sure, we don't know that, because we don't know much at all about Iberian. But what little we do know all seems to point to correspondences with Basque in names, in morphological elements, and in numerals. Now, it could be that there was massive borrowing from one direction or the other, as well as sprachbund effects. We can't disprove that. But the more parsimonious hypothesis would surely be that all the resemblances are due to the languages being related, and not even all that distantly.

- Tartessian does look different. But again, we know so little! Imagine trying to prove the existence of Indo-European, when your only resources are a grammar of modern Bengali and some fragmentary inscriptions in Old Irish! Bear in mind also that related languages can undergo substantial surface differentiation, particularly in a prehistoric context where travel may have been less widespread. (look how difficult it's proven to demonstrate large families in Amerindian).

So it's not clear to me that there WAS such diversity in Iberia.

Regarding the second point: we don't know that ANY of those languages are Neolithic in origin, let alone Mesolithic. What we do know is that by the time of our earliest attestations the region had already been invaded from at least two directions - Celts had already crossed the Pyrenees, and Greeks and Phoenicians had already arrived by sea. There is no particular reason to think that neither the Iberians (who seem related to the Aquitanians living north of the Pyrenees) nor the Tartessians (who seem anomalously developed for the area and had extensive contact with the Phoenicians and Greeks) migrated to the area not long before historic times. NON-IE doesn't necessarily mean PRE-IE.

In particular, while we could see the Basques as relicts, we could also see them as invaders. Our first depiction of them (thanks, Caesar) is not as peaceful isolated mountain dwellers in Iberia, but as a warlike race of horsemen in France. Genetically, although there is an unusually high degree of mesolithic and neolithic ancestry in the basques, they don't seem to stand to one side of the rest of europe, the way the Sardinians do, and apparently there is considerable genetic continuity across the atlantic region. We (seemingly) know that, for instance, the British Isles were invaded by the kurgans long before it is feasible to imagine the celts, per se, being there. That atlantic invasion could then be linked to, for instance, the prevalence of R1b and the Bell Beaker culture. In short, I think it's actually more attractive to see the Ur-Basques (Ur-Bell-Beakers in this hypothesis) as allies of the PIE - a precursor invasion that was then overrun by the second wave.

Even if we did see both Basque-Iberian and Tartessian as "native", in the short term, we know that Europe suffered at least two massive invasions from the middle-east with the Neolithic transition - the LBK by land, and the Cardials by sea. If these were indeed two language families (though genetically they're closely related, so that's far from certain), it wouldn't be too surprising to see two language families in Iberia, a region that could feasibly be reached by both routes (Tartessians as Cardial seafarers, Iberians as LBKs come over the Pyrenees).

Finally, regarding the idea of Iberia as representative: it seems the exact opposite to me.

For one thing, mountainous peninsulars in general aren't somewhere to look for representativeness. Even if it were true that there were 3 neolithic stocks in each peninsular, there might still only be 10-15 stocks in all of Europe, rather than 25-30. Consider how many apparently distinct families there are in California, compared to those of the great Plains... (and Iberia was much,much less diverse than California, at least by historic times).

But there are also specific reasons to doubt the representativeness of Iberia.

First, if we're talking about the Mesolithic situation, there is a big difference between the peninsular refugia and the thawing tundra. The tundra seems to have been rapidly repopulated by a single culture, but that doesn't necessarily mean that the same people took over Iberia: it would be more likely that as those people spread out from the refugium, they left whatever diversity was there to its own devices - repopulating a thawing wasteland with seminomadism is much easier than overwhelming an established fish-eating population hotspot. [not that I imagine there were many stocks surviving in that refugium anyway, but it's possible]. The Tartessians in particular are as far away from the Mesolithic expansions as possible, so they could well be only very, very distantly related to most of mesolithic europe.

Then, regarding the Neolthic invasions, Iberia is at the exact opposite end of the continent from the source of those invasions. If anywhere retained pre-neolithic languages, it would be Iberia. Likewise, Iberia would be one of the most likely places to see settlement from both directions - or indeed from a third, precursor group (settlers across the Med before the main Cardial expansion). So Iberia has particular reasons to show much more diversity than the rest of Europe - and frankly I think the interesting thing is that it doesn't show MORE diversity!


So while the crazy-quilt hypothesis is obviously not something we can currently falsify, and almost certainly never will be, it seems to me that it requires a whole series of unparsimonious assumptions, and is less likely true than untrue.
Blog: [url]http://vacuouswastrel.wordpress.com/[/url]

But the river tripped on her by and by, lapping
as though her heart was brook: Why, why, why! Weh, O weh
I'se so silly to be flowing but I no canna stay!

Zju
Lebom
Lebom
Posts: 243
Joined: Tue May 08, 2012 11:10 am

Re: European languages before Indo-European

Post by Zju »

I honestly can't tell what the argument is about. How do you define language stock?
Before the PIE hypothesis existed, Germanic, Romance, Slavi, Indic, etc. were all different stocks. After it, they became one.

It'd be useful to define what 'language stock' means first and then judge how many there were in Europe.

Ultimately, I wouldn't be surprised if all current languages descend from just 20-30 languages from 30 000 years ago, with 5-6 having given 80% of the languages today.

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

Zju wrote:I honestly can't tell what the argument is about. How do you define language stock?
Before the PIE hypothesis existed, Germanic, Romance, Slavi, Indic, etc. were all different stocks. After it, they became one.

It'd be useful to define what 'language stock' means first and then judge how many there were in Europe.

Ultimately, I wouldn't be surprised if all current languages descend from just 20-30 languages from 30 000 years ago, with 5-6 having given 80% of the languages today.
I define "language stock" as a family with a time depth in the range of 4000 to 8000 years, i.e. not obvious to the casual observer but clearly demonstrably related.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
mèþru
Smeric
Smeric
Posts: 1984
Joined: Thu Oct 29, 2015 6:44 am
Location: suburbs of Mrin
Contact:

Re: European languages before Indo-European

Post by mèþru »

In that case, it is pointless to talk about stock, because you cannot demonstrate any relationship or lack of one based off of current evidence. It's the new angels on the head of a pin.
ìtsanso, God In The Mountain, may our names inspire the deepest feelings of fear in urkos and all his ilk, for we have saved another man from his lies! I welcome back to the feast hall kal, who will never gamble again! May the eleven gods bless him!
kårroť

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

Salmoneus wrote:Nobody can know, of course.
Ignoramus et ignorabimus has always been the motto of notorious skeptics who are afraid of finding out what's going on ;) If all scholars were this kind of skeptic, nothing would ever be found out! It was one of the leading maxims of Medieval scholasticism, and one of the two main reasons why science stagnated in the Middle Ages (the other being that all scholastic endeavour was grounded in theology, which is not very productive when applied to mundane matters).
That said, I'm afraid I'm not convinced.

So far as I can see, you need three premises:
- there was linguistic diversity in Iberia
We have two patently non-IE languages and one probably non-IE language attested, and these could not yet have been shown to be related - they could be, but we don't know yet. Hence, I propose the Iberian Peninsula as a test case, as I have written in the head line. I said, and meant, test case, not proof of anything.
- that linguistic diversity preceded the kurgan invasions
I consider diversity more likely than uniformity in pre-IE Europe, or at least in Mesolithic Europe. Which factors could have led to a uniform linguistic landscape in a Mesolithic setting? (Neoliticization, if by demic diffusion - as widely assumed at least for Europe north of the Alps - may have effected in the spread of a single language family over a wide area, which may be reflected in the Old European Hydronymy if that is not a mirage of the "ley lines" type.)
- in terms of diversity, Iberia was representative of Europe as a whole
At least of Mediterranean Europe. I think diversity was less north of the Alps, and least in Scandinavia and northern Russia, each of which may have been covered by a single stock.
All three premises, in my opinion, are highly questionable.

Regarding the three families and diversity in Iberia:
- it's not even clear that Basque was present in Iberia in prehistoric times; our earliest and most voluminous attestations are from France, not Spain, and it appears that in at least part of its later Iberian territory it was overlaid on a Celtic (or similar) substrate. While it's possible that it was present all along, we can't really assume that for sure, as I understand it.
That may indeed be the case. At the time of Roman conquest, Aquitanian, i.e. a language very close to Proto-Basque, was spoken in southwestern France. The area is still called Gascony today, which of course is from Latin Vasconia. South of the Pyrenees, Proto-Basque was spoken in Navarre and east of that all the way to where now is Andorra, while the western part of Basque Country may have spoken something else, but the latter is uncertain.
- Iberian appears to be related to Basque. Sure, we don't know that, because we don't know much at all about Iberian. But what little we do know all seems to point to correspondences with Basque in names, in morphological elements, and in numerals. Now, it could be that there was massive borrowing from one direction or the other, as well as sprachbund effects. We can't disprove that. But the more parsimonious hypothesis would surely be that all the resemblances are due to the languages being related, and not even all that distantly.
Iberian may indeed be related to Basque. I never said it could not; I only said that it is not proven either way. A few Iberian name elements can, as I said, be interpreted by means of Basque, but we don't know whether these interpretations are correct and not deceptive, and many elements cannot. Similarities in phonology that seem to exist between Iberian and Proto-Basque (we don't know Iberian phonology well because the exact phonemic values of some of the Iberian letters are still controversial) may be due to a Sprachbund, and the lexical resemblances, if not bogus, may be loanwords. Yet, it could easily be the case that the languages are related.
- Tartessian does look different. But again, we know so little! Imagine trying to prove the existence of Indo-European, when your only resources are a grammar of modern Bengali and some fragmentary inscriptions in Old Irish! Bear in mind also that related languages can undergo substantial surface differentiation, particularly in a prehistoric context where travel may have been less widespread. (look how difficult it's proven to demonstrate large families in Amerindian).
Fine. When I read John T. Koch's little book, I thought, "Nice". I was not fully convinced, but the thing seemed plausible to me. Later, I read various negative reviews, and concluded that it is more likely that Tartessian was a pre-Celtic language for chronological reasons I gave in my post: the more likely date for the spread of Celtic to the southwest is at the end of the period from which we have Tartessian inscriptions, rather than at the beginning of that period. Alas, I don't know, and I cannot comment on the controversy because I am not a Celticist.
So it's not clear to me that there WAS such diversity in Iberia.
It is indeed not clear. And that is why the Iberian Peninsula is IMHO so valuable as a test case: by examining the relationship between Basque, Iberian and Tartessian, we can arrive at an estimate how many language families there were before Indo-Europeanization. If all three are related, or if Basque and Iberian are related to each other and Tartessian is IE, we have one stock. If Tartessian is IE and Basque and Iberian are unrelated to each other, or if Tartessian is non IE and two of the three languages are related at the exclusion of the third, we have two stocks. If Tartessian is non-IE and all three languages are unrelated to each other, we have three stocks.
Regarding the second point: we don't know that ANY of those languages are Neolithic in origin, let alone Mesolithic. What we do know is that by the time of our earliest attestations the region had already been invaded from at least two directions - Celts had already crossed the Pyrenees, and Greeks and Phoenicians had already arrived by sea. There is no particular reason to think that neither the Iberians (who seem related to the Aquitanians living north of the Pyrenees) nor the Tartessians (who seem anomalously developed for the area and had extensive contact with the Phoenicians and Greeks) migrated to the area not long before historic times. NON-IE doesn't necessarily mean PRE-IE.
Fair. Any of the two or three non-IE languages of the Iberian Peninsula may have arrived after the spread of Indo-European. But is that the "null hypothesis"? What speaks positively against them being autochthonous? Proposing that they are not autochthonous is the same kind of speculative scholarship you accuse me of and consider illegitimate.
In particular, while we could see the Basques as relicts, we could also see them as invaders. Our first depiction of them (thanks, Caesar) is not as peaceful isolated mountain dwellers in Iberia, but as a warlike race of horsemen in France. Genetically, although there is an unusually high degree of mesolithic and neolithic ancestry in the basques, they don't seem to stand to one side of the rest of europe, the way the Sardinians do, and apparently there is considerable genetic continuity across the atlantic region. We (seemingly) know that, for instance, the British Isles were invaded by the kurgans long before it is feasible to imagine the celts, per se, being there. That atlantic invasion could then be linked to, for instance, the prevalence of R1b and the Bell Beaker culture. In short, I think it's actually more attractive to see the Ur-Basques (Ur-Bell-Beakers in this hypothesis) as allies of the PIE - a precursor invasion that was then overrun by the second wave.
How is that less speculation than what I did in my post two days ago? Where is the evidence of "kurgan people" incursions into Britain "long before it is feasible to imagine the Celts [...] being there"? Where is the evidence of Basques as allies of the "kurgan people"? You are putting speculation against speculation.

The Bell-Beaker people seem to have originated from the Iberian Peninsula, where the oldest C14 datings have been found. Also, they seem to have been a diaspora rather than the majority population in the area where they are attested, probably merchants, as they travelled a lot (as strontium isotope analyses show) and are often found at copper, salt and other mineral production sites. That means that they probably had rather little impact on the languages of Central and Western Europe.
Even if we did see both Basque-Iberian and Tartessian as "native", in the short term, we know that Europe suffered at least two massive invasions from the middle-east with the Neolithic transition - the LBK by land, and the Cardials by sea. If these were indeed two language families (though genetically they're closely related, so that's far from certain), it wouldn't be too surprising to see two language families in Iberia, a region that could feasibly be reached by both routes (Tartessians as Cardial seafarers, Iberians as LBKs come over the Pyrenees).
Maybe; we don't know which languages these people spoke. But where are the relatives of the Paleo-Hispanic languages in Italy, on the Balkan Peninsula, in Central Europe?
Finally, regarding the idea of Iberia as representative: it seems the exact opposite to me.

For one thing, mountainous peninsulars in general aren't somewhere to look for representativeness. Even if it were true that there were 3 neolithic stocks in each peninsular, there might still only be 10-15 stocks in all of Europe, rather than 25-30. Consider how many apparently distinct families there are in California, compared to those of the great Plains... (and Iberia was much,much less diverse than California, at least by historic times).
I never claimed that all of Europe was ever as diverse as the Iberian Peninsula! You are whacking up a strawman. Iberia was indeed much less diverse than California in historic times as about half of the peninsula was Indo-Europeanized. I never claimed the Iberian Peninsula was as diverse as California. California has about a dozen indigenous stocks, the Iberian Peninsula has no evidence for any more than three. And as you (and I!) say, those three languages may be related to each other, or one of them even IE, which would mean that we have just one stock. You are battering an open door here - I have since then grown skeptical of my own high estimate of the number of pre-IE linguistic stocks in Europe, and prefer a somewhat lower number, perhaps 10 to 12.
But there are also specific reasons to doubt the representativeness of Iberia.

First, if we're talking about the Mesolithic situation, there is a big difference between the peninsular refugia and the thawing tundra. The tundra seems to have been rapidly repopulated by a single culture,
Which not necessarily spoke a single language; compare the famous mismatch of language families and cultural areas in indigenous North America. But surely, the linguisitic diversity in the Mesolithic Central Zone (between the Alpide mountain system and the North and Baltic Seas) was more like the Great Plains than California, and lower than in the Iberian Peninsula. But note that the Central Zone was not devoid of human inhabitants during the last ice age; in fact, it was rich in big game to hunt. Surely, more people moved in from the south when the ice age ended, the climate zones shifted and people followed their game northwards. So we have at least two linguistic layers - and at least two language families - in the Central Zone.

The only parts of Europe which seem to have been uninhabited in the last ice age were those which were actually covered with ice; these would have very low linguistic diversity in the Mesolithic as they had been freshly populated for the first time. There may have been a single family of languages all the way from Ireland to the Kola Peninsula. But as the area probably was peopled from more than one source, there may just as well have been two or three. All of them extensions of Central Zone families.
but that doesn't necessarily mean that the same people took over Iberia: it would be more likely that as those people spread out from the refugium, they left whatever diversity was there to its own devices - repopulating a thawing wasteland with seminomadism is much easier than overwhelming an established fish-eating population hotspot. [not that I imagine there were many stocks surviving in that refugium anyway, but it's possible]. The Tartessians in particular are as far away from the Mesolithic expansions as possible, so they could well be only very, very distantly related to most of mesolithic europe.
Yes, the Tartessians were far away from the movements within Europe, but close to the Straits of Gibraltar, which certainly have always seen migrations from Africa to Europe and vice versa. It is not out of the question that Tartessian originated in Norhwestern Africa. The same could of course also be true for Iberian.
Then, regarding the Neolthic invasions, Iberia is at the exact opposite end of the continent from the source of those invasions. If anywhere retained pre-neolithic languages, it would be Iberia. Likewise, Iberia would be one of the most likely places to see settlement from both directions - or indeed from a third, precursor group (settlers across the Med before the main Cardial expansion). So Iberia has particular reasons to show much more diversity than the rest of Europe - and frankly I think the interesting thing is that it doesn't show MORE diversity!
You are right that the Iberian Peninsula probably was more diverse than most other parts of Europe. At least more diverse than France, or Central Europe. Italy is smaller and therefore probably had fewer autochthonous languages (and there are reasons, though not all too forceful, to suspect that Etruscan may have come from NW Anatolia around 1200 BC); the Balkan Peninsula is of similar size but closer to the origins of Neolithic migrations.
So while the crazy-quilt hypothesis is obviously not something we can currently falsify, and almost certainly never will be, it seems to me that it requires a whole series of unparsimonious assumptions, and is less likely true than untrue.
It may be more parsimonious to assume that pre-IE Europe was linguistically uniform, but doesn't this version of Ockham's Razor cut the skin? Sure, there probably weren't as many as 20 or even 30 stocks in Mesolithic Europe, and we may not have evidence for more than half a dozen on all three Mediterranean peninsulas put together, but I frankly don't see the forces that could have led to continent-wide language families before the Neolithic.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

Travis B.
Sumerul
Sumerul
Posts: 3570
Joined: Mon Jun 20, 2005 12:47 pm
Location: Milwaukee, US

Re: European languages before Indo-European

Post by Travis B. »

I like "Wir müssen wissen — wir werden wissen!" better as a slogan myself... even though of course a number of Hilbert's questions have been found to be unanswerable.
Dibotahamdn duthma jallni agaynni ra hgitn lakrhmi.
Amuhawr jalla vowa vta hlakrhi hdm duthmi xaja.
Irdro. Irdro. Irdro. Irdro. Irdro. Irdro. Irdro.

User avatar
Zaarin
Smeric
Smeric
Posts: 1136
Joined: Sun Aug 15, 2010 5:00 pm

Re: European languages before Indo-European

Post by Zaarin »

I have no horse in this race, but I would contribute a couple comments:
science stagnated in the Middle Ages
This Enlightenment notion has long since been refuted. Sure, the advancement of science slowed in the wake of Rome's collapse and the ensuing widespread illiteracy, but the High and Late Middle Ages were a period of significant advancement and dedicated scholarship (not to mention a flowering of arts and literature). The Renaissance would not have even been possible without its predecessor, the Twelfth Century Renaissance. Scholasticism may not appear advanced compared to modern science, but it was an important step in the pathway to the development of modern scientific methods.
Where is the evidence of "kurgan people" incursions into Britain "long before it is feasible to imagine the Celts [...] being there"?
The Bell Beaker people of Bronze Age Britain were almost certainly Indo-European and definitely not Celtic, at least not Celts as we know them post-Halstatt--perhaps we could call them "pre-Celts." I have nothing to contribute about where the Beaker culture originated--I'm most familiar with its presence in Britain and France--but they do strongly suggest that Indo-European people were in Britain long before the Celts.
"But if of ships I now should sing, what ship would come to me,
What ship would bear me ever back across so wide a Sea?”

User avatar
Salmoneus
Sanno
Sanno
Posts: 3197
Joined: Thu Jan 15, 2004 5:00 pm
Location: One of the dark places of the world

Re: European languages before Indo-European

Post by Salmoneus »

Science also does not move forward much if we pick a pet theory and proclaim it regardless of evidence, decrying all criticism as an attack on the very principles of Western Civilisation. But, each to their own. Except to note that I was actually offering "we can never know" as a defence of your improbable hypothesis...

My general argument would simply be:

- time and time again, when we see rapid, massive expansion by a particular, homogenous genetic and cultural group, we also see areas dominated by large language families, rather than by a patchwork of local isolates and small families;

- Europe's history has been dominated by a series of exactly this sort of rapid, massive expansions by particular genetically and culturally homogenous groups

- therefore it is most likely that Europe's history has also been dominated by large language families.

It is, of course, possible that things were less uniform than they appear. It is certain possible to have languages spoken for unfathomable reason, despite the population histories of their speakers (although I'm not aware of any specific examples - the opposite (where there is more linguistic homogeneity than expected) is much more common (eg the Sami)).]

However, such a claim, in my view, requires evidence.
I said, and meant, test case
Perhaps this is a translation issue. Because from your other comments, I do not see what you think you are testing or how it comprises a test. I assumed that you were using neolithic diversity in iberia to test the hypothesis of neolithic diversity outside iberia. But apparently not. If you were just testing whether between zero and three non-IE language families are attested in Iberia, then yes, great, consider that tested, but I'm not sure such a narrow claim is particularly interesting.
Which factors could have led to a uniform linguistic landscape in a Mesolithic setting?
The fact that the continent had been conquered by a genetically and culturally homogenous group that largely replaced the previous inhabitants. Languages can spread without massive demic replacement, but it's much less likely for languages to fail to spread even when massive demic replacement occurs. Not impossible, no, but not the default case. Certainly not when you want it to have failed to happen again and again and again everywhere throughout the continent. Maybe instead the new group comprised a number of different, unrelated linguistic groups, despite their shared genes and culture? Maybe. But again, this is not the parsimonious hypothesis in my opinon.
At least of Mediterranean Europe. I think diversity was less north of the Alps, and least in Scandinavia and northern Russia, each of which may have been covered by a single stock.
OK, that's certainly more reasonable.
South of the Pyrenees, Proto-Basque was spoken in Navarre and east of that all the way to where now is Andorra
We don't actually know that, though, do we? I'm not sure there are any actual attestations. Our evidence is just that the Roman geographers thought that people called 'Vascones' lived there, and we assume that because this sounds like 'Basque' the Vascones must have spoken proto-Basque. But then, if we're trusting Roman geographers so much, we might also want to listen when they say that Iberians and Vascones spoken related languages...
A few Iberian name elements can, as I said, be interpreted by means of Basque, but we don't know whether these interpretations are correct
Ah, so you're happy to resort to "skepticism" as you call it, in order to defend your theories!
In reality we do not know that they were not related. Therefore their unrelatedness cannot be used to "test" anything...
Later, I read various negative reviews, and concluded that it is more likely that Tartessian was a pre-Celtic language
To be clear, I also assume that Tartessian was presumably not Celtic (or IE). I was drawing an analogy between the difficulty of demonstrating an IE family from limited data and the presumed difficulty of demonstrating a putative Tartesso-Basque (let alone Tartesso-Etruscan or Tartesso-Minoan or the like). Again, it is not clear that Iberian and Tartessan were not related - if they were related in any but the most superficially obvious way, we probably would still not have discovered this, so that possibility cannot be discounted. So again, our "test" shows only that there was at least one non-IE language spoken in Iberia after the IE invasions. I'm not sure what that test result is meant to imply.

And that is why the Iberian Peninsula is IMHO so valuable as a test case: by examining the relationship between Basque, Iberian and Tartessian, we can arrive at an estimate how many language families there were before Indo-Europeanization.
a) what is it testing?
b) no, we can't, because our access to data about these languages is extremely scant (and in my case also because I'm not a historical linguistics genius, but I don't want to speak for your own prowess in proving the world wrong; maybe you are indeed the next Ventriss)
c) no, we can't, because we know nothing about how many language families there were before Indo-Europeanization.

Consider three scenarios:
1. Iberia is inhabited by 32 language families. Indo-Europeans invade and overwhelm most of the peninsular. IE technologies are passed to some native groups, who take advantage of the weakening of other groups to assimilate some of their neighbours. We end up with IE and between 1 and 3 surviving pre-IE families.

2. Iberia is inhabited by between 1 and 3 native families, covering the whole of the peninsular. Indo-Europeans invade and overwhelm most of the peninsular. The non-IE families have their ranges drastically reduced in scope, but nonetheless survive in some areas. We end up with IE and between 1 and 3 surviving pre-IE families.

3. Iberia is inhabited by between 1 and 32 native families. Indo-Europeans and proto-Vasco-Iberians (and/orTartessians) invade and overwhelm most of the peninsular. Almost all native families are wiped out, leaving only Tartessian, or possibly Vasconic (or Iberian, if not Vasconic) as the only survivor. Or there are no survivors. We end up with IE and between 0 and 3 pre-IE families.

All three of these scenarios are eminently plausible. Well, maybe not 32 language families precisely, but a dozen or more could easily fit compare California). Indeed, this is probably more likely that imagining that all the pre-IE languages survived into historic times - there's no particular reason to imagine Tartessian extending up to Galicia, or Iberia extending across to the coast of Portugal, after all. On the other hand, that's also certainly a possibility. And on the third hand, given the general chaos of the era and the fact that there's between 1000 and 500 years between the time of the conquest of most of Europe by IEs and the time of the attestation of Tartessian - and another half-millennium or more before Iberian and then Aquitanian enter the scene! - it's far from implausible that they could have been other migrations into the area from the north, the east or the south and that some or all of the attested palaeo-iberian languages are not in fact Old European in origin.

So this "test" tells us nothing, unless I've entirely failed to understand what you think you're testing and how this test is to be applied.
Proposing that they are not autochthonous is the same kind of speculative scholarship you accuse me of and consider illegitimate./quote] No - assuming that they are not autochthonous and using the "result" of that "test" as evidence of a wider hypothesis about Mesolithic Europe (which, let's not forget, was thousands of years and a massive population replacement prior to the era we have evidence for) would be the kind of speculative "scholarship" that I find tendentious...
How is that less speculation than what I did in my post two days ago? Where is the evidence of "kurgan people" incursions into Britain "long before it is feasible to imagine the Celts [...] being there"? Where is the evidence of Basques as allies of the "kurgan people"? You are putting speculation against speculation.
Of course it's speculation. And it's entirely legitimate to put speculation against speculation. In a trial, if the prosecution suggests a theory of what happened, it's entirely legitimate for the defence to offer an alternative plausible theory. The whole point is that if both theories are plausible, you can't in good conscience convict. The defence doesn't have to prove that the defendant didn't do it, only that there exists reasonable doubt.

That said, the bit you focus on is not particularly speculative. We know that that the neolithic population of Britain and Ireland was almost entirely replaced by a completely different, "Bell Beaker" population from central or eastern Europe. We know that this new population was essentially genetically indistinguishable from the later Indo-European population of the islands. We also know, however, that this occured at a time when it seems implausible to imagine Celts being in the Isles. Archeologically, the Celts are assumed at that time to have been in central or eastern Europe. Linguistically, Celtic does not look sufficiently distinct from other IE branches to have split off at such an early time; nor do the surviving Celtic languages look sufficiently distinct from one another (relative to their demonstrable changes since the split!) to make it attractive to see their original differentiation as having occured so early (though it would still be possible to, say, suggest two celtic migrations into ireland).
So, we're left with a non-Celtic steppe invasion of the Isles. We could call these proto-Celts, but then there's no actual reason to do so (rather than calling them proto-Italics or proto-Germans or whatever). We could call them an unattested IE branch. We could call them a sister of IE that also partook in the migrations. Or we could call them a neighbouring tribe who went along with the PIEs when they invaded but originally spoke an unrelated (or distantly related) language.
Again, these hypotheses are all plausible, so we can't really say which one is true.

However, since we can say that there was a language family in Atlantic Europe with no known children, and since we can say that there is now a language family in Atlantic Europe with no known parent, and since the genetics of the two groups seem similar, and since the earliest cultural data we have for the later family makes them look not dissimilar from the presumed culture of the former family, it is at least an appealing hypothesis that the later family is actually a descendent of the former. But sure, it's speculative. There are reasons to like the idea (it explains neatly why the Basques look like Bell Beakers - because they are - and doesn't require the Basques to be two different survivals at the same time (a genetic survival of the BBs and also a linguistic survival of the pre-BBs)), and reasons not to like the idea (it requires us to believe that the kurgans were at least two genetically and culturally homogenous but linguistically seemingly highly distinct groups, which is certainly possible but not the simplest hypothesis).

[Now, if those Vasco-Caucasian crackpots could prove their case, that'd make it very interesting...]
The Bell-Beaker people seem to have originated from the Iberian Peninsula, where the oldest C14 datings have been found. Also, they seem to have been a diaspora rather than the majority population in the area where they are attested, probably merchants, as they travelled a lot (as strontium isotope analyses show) and are often found at copper, salt and other mineral production sites. That means that they probably had rather little impact on the languages of Central and Western Europe.
It's true that the earliest physical remains are from the Atlantic. However, genetically they appear to have originated from central or eastern Europe, and to have completely replaced the majority population in the area where they are attested.

You could, of course, say that these are two different groups - that a migrant group from the east and a pottery group from the west just happened to have coinciding areas. That's possible. Although I'm not aware of any reason to think that your 'isolated merchant caste' hypothesis would be plausible, rather than plain spread of culture into a new group.
But where are the relatives of the Paleo-Hispanic languages in Italy, on the Balkan Peninsula, in Central Europe?
Gone, obviously. These areas have all been bulldozed by IE, so we can't legitimately say what was or wasn't there before then. Of course, Etruscan could have been related to Palaeo-Hispanic languages (given the lack of knowledge, a distant relation wouldn't be readily apparent). Minoan or Eteocypriot (if they're not related to one another) might be related to Palaeo-Hispanic - we couldn't possibly know, until a lot more success has been had with those scripts. And of course a pre-IE language may have survived on Sardinia right the way into the historical era, but no attestations are known. Absence of evidence, however, is not evidence of absence.
I never claimed that all of Europe was ever as diverse as the Iberian Peninsula!
Then, again, what on earth are you "testing"? If you accept that the situation in Iberia tells us nothing at all about the situation outside of Iberia, how can Iberia be a "test case" of any theory?
But note that the Central Zone was not devoid of human inhabitants during the last ice age; in fact, it was rich in big game to hunt. Surely, more people moved in from the south when the ice age ended, the climate zones shifted and people followed their game northwards. So we have at least two linguistic layers - and at least two language families - in the Central Zone.
There have actually been several phases of replacement. As I understand it:

There were several phases before the glacial maximum. The earliest phases appear completely genetically unrelated to modern Europeans (more accurately: they are no more related to Europeans than to East Asians). Before the ice, Europe appears to have been genetically diverse.
When the ice came, the tundra region became more or less entirely depopulated, and the groups immediately preceding the ice do not appear to have had any genetic continuity with later groups.
As the ice receded, the big game hunters entered the region in much larger numbers. If there were any inhabitant before them, they became obsolete, leaving no apparent trace. The big game hunters were from Iberia (or southern France), and were related to one of the earlier groups, but not closely related to their immediate predecessors in the north.
The big game hunters brought about complete homogeneity in Europe, replacing the earlier diversity. At this point, remains everywhere from Spain to Hungary (although I don't know about the Balkans) became effectively genetically indistinguishable. The same culture was also rapidly spread over most of this area. It seems extremely plausible to me that this invasion spread a single language family, or at least no more than a couple, across the whole continent.

Later, however, at the transition into the Mesolithic, another group arose. This group had some sort of connection to the Middle East, although it's too early to say whether they were immigrants, or emigrants, or just cousins. This new group rapidly spread throughout the whole continent; the earlier group left some traces, but were clearly genetically overwhelmed. Again, this new group brought with them their own distinctive culture. After this latest invasion and population replacement, Europe was again culturally and genetically homogenous. Again, it seems reasonable to suspect that it was also linguistically homogenous, or nearly so.

A few thousand years later, two big invasions occured from the middle east, associated with the introduction of farming. Although there were two invasions, the two groups seem genetically very close together. The earlier Mesolithic groups were marginalised; they left a substantial trace in most areas, but were clearly overwhelmed and replaced. [for one thing, Mesolithic Europeans were black (by modern standards), and Neolithic Europeans were not]. So again, after these invasions, Europe was genetically and (in two parts) culturally homogenous. It's not unreasonable to suspect that, perhaps outside of a few isolated Mesolithic holdout areas, it was also linguistically homogenous.

Partway through the Neolithic, it seems like bad things happened. It was once thought that perhaps there was another invasion at this point. People don't think that anymore, I don't think - at least, if there were invasions, they were not massive population replacements. Maybe there were raiders? Probably more likely it's just that as the farmers spread out, they lost their cohesion and began to fall into conflict. From this point on, the genetic influence of the Mesolithic actually increases a little - some have suggested a "fightback", but it seems more likely this was just the gradual assimilation of the last holdouts, and that once they were assimilated they were actually sexually quite succesful.

Then metalworking comes along. Some time later, there was a massive invasion from the Steppe by the Yamna (or a related group), who were partly distantly related to the mesolithics and partly not. Although their influence is less extreme in western and southern Europe (particularly in the Basque Country, where there is an unusally high level of mesolithic ancestry), the Yamna (or "kurgans") effectively overwhelmed the entire continent (apart from Sardinia). Again, Europe became fairly culturally and genetically homogenous (although not to the extent of earlier invasions). And Europe also became linguistically homogenous, so far as we can tell (and eventually it became almost entirely linguistically homogenous). This doesn't prove, of course, that the more complete earlier population turnovers didn't produce less linguistic domination, but it does seem, on the face of it, to strengthen the case somewhat...
Blog: [url]http://vacuouswastrel.wordpress.com/[/url]

But the river tripped on her by and by, lapping
as though her heart was brook: Why, why, why! Weh, O weh
I'se so silly to be flowing but I no canna stay!

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

Salmoneus wrote:Science also does not move forward much if we pick a pet theory and proclaim it regardless of evidence, decrying all criticism as an attack on the very principles of Western Civilisation. But, each to their own. Except to note that I was actually offering "we can never know" as a defence of your improbable hypothesis...
One moment. You may think I am a crackpot like Octaviano, and I admit that my hypothesis is speculative. Also, I should not have drawn medieval scholasticism into this discussion; I apologize for that. But what I state is just a working hypothesis which needs to be tested against the available evidence. But we are getting into a metascientific debate here.
My general argument would simply be:

- time and time again, when we see rapid, massive expansion by a particular, homogenous genetic and cultural group, we also see areas dominated by large language families, rather than by a patchwork of local isolates and small families;
Certainly, such expansions lead to the establishment of large language families - I never claimed the opposite. At least two such expansions have affected the Iberian Peninsula: those of Celtic and Latin. There certainly were earlier ones. Yet, sometimes languages from earlier strata survive in remote pockets, which seems to have been the case with Basque.
- Europe's history has been dominated by a series of exactly this sort of rapid, massive expansions by particular genetically and culturally homogenous groups

- therefore it is most likely that Europe's history has also been dominated by large language families.
Sure. But when half a dozen expansions of this kind happen in the same area one after the other, and in each case, remnants of the earlier landscape survive in residual zones such as mountain ranges or islands, the result can be quite some diversity. I don't speak of dozens of language families in the Iberian Peninsula, but of one, two or three - dependent on whether the ancient peninsular languages are related to each other or not. And as you say, one, two or all three of those languages may have come rather lately from somewhere else.
It is, of course, possible that things were less uniform than they appear. It is certain possible to have languages spoken for unfathomable reason, despite the population histories of their speakers (although I'm not aware of any specific examples - the opposite (where there is more linguistic homogeneity than expected) is much more common (eg the Sami)).

However, such a claim, in my view, requires evidence.
Sure. As long as we don't know which of the three non-IE languages of the Iberian Peninsula is related to which, we don't know how many families they represent - that's a truism.
I said, and meant, test case
Perhaps this is a translation issue. Because from your other comments, I do not see what you think you are testing or how it comprises a test. I assumed that you were using neolithic diversity in iberia to test the hypothesis of neolithic diversity outside iberia. But apparently not. If you were just testing whether between zero and three non-IE language families are attested in Iberia, then yes, great, consider that tested, but I'm not sure such a narrow claim is particularly interesting.
Maybe I got the meaning of the English expression "test case" wrong. As you know, English is not my native language. The Neolithic linguistic diversity outside the Iberian Peninsula cannot be tested by comparing the three non-IE languages of the Iberian Peninsula in order to determine how many stocks they belong to. We would get some result about the minimum number of stocks in the Iberian Peninsula, which tells us nothing to rely on about the numbers of stocks elsewhere.
Which factors could have led to a uniform linguistic landscape in a Mesolithic setting?
The fact that the continent had been conquered by a genetically and culturally homogenous group that largely replaced the previous inhabitants. Languages can spread without massive demic replacement, but it's much less likely for languages to fail to spread even when massive demic replacement occurs.
Sure. But how many such expansions happened before the Neolithic? We don't know. There of course was one - when Homo sapiens entered the continent. But more than 30,000 years later, that stock would have disintegrated long ago. Note that by "stock" I mean a certain range of time depth - 4,000 to 8,000 years. Things like Indo-European or Uralic are today. Another large expansion probably happened after the end of the last ice age, when Northern Europe was peopled for the first time as the ice receded. There thus were no new stocks in the North, only extensions of stocks already existing in the central zone. Also, people probably moved from the Mediterranean peninsulas into the central zone but not necessarily obliterated all the languages that already were there. Note that, as I wrote earlier today, the central zone was not devoid of human habitation before the end of the ice age!
Not impossible, no, but not the default case. Certainly not when you want it to have failed to happen again and again and again everywhere throughout the continent. Maybe instead the new group comprised a number of different, unrelated linguistic groups, despite their shared genes and culture? Maybe. But again, this is not the parsimonious hypothesis in my opinon.
I don't see how the hypothesis that there have been numerous demic diffusions that homogenized the linguistic landscape of Europe over and over again without leaving residues of the old landscape in residual zones is more parsimonious than the hypothesis that the languages developed mostly in situ and old stocks broke up more and more. But almost certainly, the truth lies somewhere in between. There certainly were significant migrations and language expansions between the first peopling of Europe by our species and the Neolithicization, but I doubt that any of these affected all of Europe and removed all old languages. The northward movements at the end of the ice age as the climate zones shifted and the animals they hunted followed the shifting climate zones probably were the greatest event of this kind that occurred in the meantime.
At least of Mediterranean Europe. I think diversity was less north of the Alps, and least in Scandinavia and northern Russia, each of which may have been covered by a single stock.
OK, that's certainly more reasonable.
Fine.
South of the Pyrenees, Proto-Basque was spoken in Navarre and east of that all the way to where now is Andorra
We don't actually know that, though, do we? I'm not sure there are any actual attestations. Our evidence is just that the Roman geographers thought that people called 'Vascones' lived there, and we assume that because this sounds like 'Basque' the Vascones must have spoken proto-Basque. But then, if we're trusting Roman geographers so much, we might also want to listen when they say that Iberians and Vascones spoken related languages...
This is the first time I hear someone doubt that the Vascones of the Roman Era spoke anything like Basque! The Aquitanian language is only known from names occurring in Latin inscriptions of the Roman Era, but these names are, as I have been informed (R. L. Trask, The History of Basque, 1997), obviously Proto-Basque. Yet, I am no Vascologist and cannot evaluate whether Trask was right or wrong, but he was a major authority on the language. Is your knowledge of the relevant facts better than Trask's?
A few Iberian name elements can, as I said, be interpreted by means of Basque, but we don't know whether these interpretations are correct
Ah, so you're happy to resort to "skepticism" as you call it, in order to defend your theories!
In reality we do not know that they were not related. Therefore their unrelatedness cannot be used to "test" anything...
I never said that we know that Basque and Iberian are unrelated. Stop whacking strawmen all the time! Of course we don't know yet. Indeed, there is some evidence of relationship, but not enough to prove the relationship, and other explanations of the evidence must be considered.
Later, I read various negative reviews, and concluded that it is more likely that Tartessian was a pre-Celtic language
To be clear, I also assume that Tartessian was presumably not Celtic (or IE).
OK. So we agree on this point, though there is no proof yet either way.
I was drawing an analogy between the difficulty of demonstrating an IE family from limited data and the presumed difficulty of demonstrating a putative Tartesso-Basque (let alone Tartesso-Etruscan or Tartesso-Minoan or the like). Again, it is not clear that Iberian and Tartessan were not related - if they were related in any but the most superficially obvious way, we probably would still not have discovered this, so that possibility cannot be discounted. So again, our "test" shows only that there was at least one non-IE language spoken in Iberia after the IE invasions. I'm not sure what that test result is meant to imply.
As I have clarified above, by "stock" I don't mean any kind of relationship however remote, but a family in a specific range of time depth, namely 4,000 to 8,000 years. Things like Nostratic or Dene-Caucasian, even if they were real, wouldn't be "stocks" because their time depths would fall far outside that range. It seems that we have been talking past each other here.
And that is why the Iberian Peninsula is IMHO so valuable as a test case: by examining the relationship between Basque, Iberian and Tartessian, we can arrive at an estimate how many language families there were before Indo-Europeanization.
a) what is it testing?
b) no, we can't, because our access to data about these languages is extremely scant (and in my case also because I'm not a historical linguistics genius, but I don't want to speak for your own prowess in proving the world wrong; maybe you are indeed the next Ventriss)
I don't seriously expect to be the next Ventris; I just like to think about these matters, do some research with my limited resources and contribute to the discussion. But I hope to be more like Ventris than like Octaviano. The difference is that Ventris kept in touch with other scholars in the field, and was ready to abandon a hypothesis that yielded no useful results, while Octaviano stubbornly clung to his ideas and, in doing so, went as far as to declare handbook knowledge wrong.
c) no, we can't, because we know nothing about how many language families there were before Indo-Europeanization.
As I said above, "test case" may have been a misnomer on my side. All I meant was that by examining the relationships between the Paleo-Hispanic languages, one can get an idea of how many stocks there were, which could be helpful in guiding estimations of the linguistic diversity at least of the other two Mediterranean peninsulas.
Consider three scenarios:
1. Iberia is inhabited by 32 language families. Indo-Europeans invade and overwhelm most of the peninsular. IE technologies are passed to some native groups, who take advantage of the weakening of other groups to assimilate some of their neighbours. We end up with IE and between 1 and 3 surviving pre-IE families.

2. Iberia is inhabited by between 1 and 3 native families, covering the whole of the peninsular. Indo-Europeans invade and overwhelm most of the peninsular. The non-IE families have their ranges drastically reduced in scope, but nonetheless survive in some areas. We end up with IE and between 1 and 3 surviving pre-IE families.

3. Iberia is inhabited by between 1 and 32 native families. Indo-Europeans and proto-Vasco-Iberians (and/orTartessians) invade and overwhelm most of the peninsular. Almost all native families are wiped out, leaving only Tartessian, or possibly Vasconic (or Iberian, if not Vasconic) as the only survivor. Or there are no survivors. We end up with IE and between 0 and 3 pre-IE families.

All three of these scenarios are eminently plausible. Well, maybe not 32 language families precisely, but a dozen or more could easily fit compare California).
I see what you are getting at. Indeed, many scenarios are possible, and the number of stocks resulting from examinations of the relationships between the attested Paleo-Hispanic languages may have little to do with the actual diversity of the pre-IE Iberian Peninsula because a) entire stocks could have disappeared completely and b) some or all of the attested non-IE languages may have moved in later. But I don't see how b) could be the "null hypothesis"; and addressing a), the "null hypothesis" would be that perhaps one or two further stocks were eradicated (in the northwestern part of the peninsula where only IE languages are attested), or if the three attested non-IE languages happen to turn out to belong to one stock, it would seem likely that the rest of the peninsula was covered by the same stock as well.
Indeed, this is probably more likely that imagining that all the pre-IE languages survived into historic times - there's no particular reason to imagine Tartessian extending up to Galicia, or Iberia extending across to the coast of Portugal, after all.
Certainly, the northwest of the peninsula was occupied by languages that were different from those we find attested in the south and east - no Tartessian in Galicia or Iberian in Portugal. But were these lost languages related to Basque, Iberian and/or Tartessian or not? We simply do not know.
On the other hand, that's also certainly a possibility. And on the third hand, given the general chaos of the era and the fact that there's between 1000 and 500 years between the time of the conquest of most of Europe by IEs and the time of the attestation of Tartessian - and another half-millennium or more before Iberian and then Aquitanian enter the scene! - it's far from implausible that they could have been other migrations into the area from the north, the east or the south and that some or all of the attested palaeo-iberian languages are not in fact Old European in origin.
Your objections are valid. The languages are attested from different time periods. Tartessian ceased to be written some time in the 6th century BC, which I conjecture to be the time the region became Celtic-speaking. The Iberians took over their script from the Tartessians, but the community remained intact until about 200 BC when the Romans conquered them. Aquitanian is only known from the Roman Era, which began, in Aquitania, around 50 BC. Basque proper is known only from the Late Medieval onwards, though Mitxelena managed - by applying the comparative method to dialects and Latin loanwords (comparison between their modern forms and their known Latin antecedents allows to retrace the sound changes the language underwent) as well as internal reconstruction - to reconstruct what the Basque language looked like 2,000 years ago, and this is in broad agreement with the Aquitanian evidence.
So this "test" tells us nothing, unless I've entirely failed to understand what you think you're testing and how this test is to be applied.
OK, so not a "test". I apologize for this misnomer. I don't know what to call it, a "probe" or a "sample"? My idea was simply to get an idea of the linguistic diversity of a Mediterranean peninsula by comparing the three non-IE Paleo-Hispanic languages.
Proposing that they are not autochthonous is the same kind of speculative scholarship you accuse me of and consider illegitimate.
No - assuming that they are not autochthonous and using the "result" of that "test" as evidence of a wider hypothesis about Mesolithic Europe (which, let's not forget, was thousands of years and a massive population replacement prior to the era we have evidence for) would be the kind of speculative "scholarship" that I find tendentious...
Surely, a lot of things happened between the Mesolithic and the time when the people of the ancient Iberian Peninsula started writing their languages, and it is hard, if not impossible, to reconstruct what was going on.
How is that less speculation than what I did in my post two days ago? Where is the evidence of "kurgan people" incursions into Britain "long before it is feasible to imagine the Celts [...] being there"? Where is the evidence of Basques as allies of the "kurgan people"? You are putting speculation against speculation.
Of course it's speculation. And it's entirely legitimate to put speculation against speculation. In a trial, if the prosecution suggests a theory of what happened, it's entirely legitimate for the defence to offer an alternative plausible theory. The whole point is that if both theories are plausible, you can't in good conscience convict. The defence doesn't have to prove that the defendant didn't do it, only that there exists reasonable doubt.
Of course. Note again that I am not convinced that my ideas are right. They could be utter hogwash.
That said, the bit you focus on is not particularly speculative. We know that that the neolithic population of Britain and Ireland was almost entirely replaced by a completely different, "Bell Beaker" population from central or eastern Europe. We know that this new population was essentially genetically indistinguishable from the later Indo-European population of the islands. We also know, however, that this occured at a time when it seems implausible to imagine Celts being in the Isles.
What you say is actually quite in accordance with my own hypothesis of the languages of the Bronze Age British Isles, which in my hypothesis were not Celtic but part of a sister group of Indo-European I call "Aquan" - the language of the Old European Hydronymy. I am not sure of the Bell Beaker connection I proposed earlier, but grew doubtful of for the reasons I have already stated in this thread. But if the Bell Beaker people originate from central or eastern Europe, are genetically related to Indo-Europeans and were more than just a diaspora, that's perfectly fine from the standpoint of the Aquan hypothesis.
Archeologically, the Celts are assumed at that time to have been in central or eastern Europe. Linguistically, Celtic does not look sufficiently distinct from other IE branches to have split off at such an early time; nor do the surviving Celtic languages look sufficiently distinct from one another (relative to their demonstrable changes since the split!) to make it attractive to see their original differentiation as having occured so early (though it would still be possible to, say, suggest two celtic migrations into ireland).
On this point, I whole-heartedly agree with you. Celtic, unlike Anatolian, doesn't look like a particularly early breakaway from the rest of IE, and the degree of diversity of the attested Celtic languages points at a Proto-Celtic no earlier than about 1000 BC, probably to be associated with the Hallstatt culture of Late Bronze Age/Early Iron Age southern central Europe. (I fancy the "P-Celtic" shift *kw > *p to have spread through the Celtic dialect continuum as a fashionable shibboleth of the more sophisticated La Tène culture, failing to reach the outliers in Ireland and the Iberian Peninsula.)
So, we're left with a non-Celtic steppe invasion of the Isles. We could call these proto-Celts, but then there's no actual reason to do so (rather than calling them proto-Italics or proto-Germans or whatever). We could call them an unattested IE branch. We could call them a sister of IE that also partook in the migrations. Or we could call them a neighbouring tribe who went along with the PIEs when they invaded but originally spoke an unrelated (or distantly related) language.
Again, these hypotheses are all plausible, so we can't really say which one is true.
Fair. This is also what I think of the matter. And I am not alone. Francisco Villar said something similar in his book Los Indoeuropeos y los orígines de Europa - he considers the language of the Old European Hydronymy to be an early breakaway from PIE that was later clobbered by the historically known IE languages. That language would be a centum language (thus not Balto-Slavic) which merged PIE *o into *a (thus not Italo-Celtic) and did not shift its stops (thus not Germanic; the last point seems uncertain to me, though I am sure for other reasons I shall immediately come to that it wasn't Germanic). In my opinion, the dominant /a/-vocalism in the OEH is a problem here. Villar's idea raises the question: Why is everything in o-grade? To me, it looks like a language group that broke off before the rise of the PIE ablaut system, and preserved the pre-ablaut 3-vowel system. My ongoing work on a substratum dictionary seems to corroborates this hypothesis.

Alas, all that is speculation and I am not sure that I am right.
However, since we can say that there was a language family in Atlantic Europe with no known children, and since we can say that there is now a language family in Atlantic Europe with no known parent, and since the genetics of the two groups seem similar, and since the earliest cultural data we have for the later family makes them look not dissimilar from the presumed culture of the former family, it is at least an appealing hypothesis that the later family is actually a descendent of the former.
Fair. It is not unlikely that Basque once had relatives that were spoken in a larger area; yet, Vennemann's evidence for a Vasconic family extending as far east as Poland and the Baltic countries is not very convincing.
But sure, it's speculative. There are reasons to like the idea (it explains neatly why the Basques look like Bell Beakers - because they are - and doesn't require the Basques to be two different survivals at the same time (a genetic survival of the BBs and also a linguistic survival of the pre-BBs)), and reasons not to like the idea (it requires us to believe that the kurgans were at least two genetically and culturally homogenous but linguistically seemingly highly distinct groups, which is certainly possible but not the simplest hypothesis).
An interesting idea - Basque coming from the east together with Indo-European. But why, then, don't we find other Vasconic languages elsewhere in Europe?
[Now, if those Vasco-Caucasian crackpots could prove their case, that'd make it very interesting...]
Actually, Vasco-Caucasian may be correct - only the evidence adduced by people like Octaviano is shoddy. Basque and the two North Caucasian families may be the last remains of the same old layer of languages, alas, there seems to be no way knowing, and the hypothesis is based on superficial resemblances such as ergativity which prove nothing.
The Bell-Beaker people seem to have originated from the Iberian Peninsula, where the oldest C14 datings have been found. Also, they seem to have been a diaspora rather than the majority population in the area where they are attested, probably merchants, as they travelled a lot (as strontium isotope analyses show) and are often found at copper, salt and other mineral production sites. That means that they probably had rather little impact on the languages of Central and Western Europe.
It's true that the earliest physical remains are from the Atlantic. However, genetically they appear to have originated from central or eastern Europe, and to have completely replaced the majority population in the area where they are attested.

You could, of course, say that these are two different groups - that a migrant group from the east and a pottery group from the west just happened to have coinciding areas. That's possible. Although I'm not aware of any reason to think that your 'isolated merchant caste' hypothesis would be plausible, rather than plain spread of culture into a new group.
Now that I think of it it makes perfect sense. There may be a merchant caste, and it may have originated in the Iberian Peninsula, but they may be of a racial stock that spread from the east earlier. This also explains why their range is roughly coterminous with the Old European Hydronymy - they travelled where languages closely related to their own language were spoken such that they could make themselves understood among the locals, and no farther.
But where are the relatives of the Paleo-Hispanic languages in Italy, on the Balkan Peninsula, in Central Europe?
Gone, obviously. These areas have all been bulldozed by IE, so we can't legitimately say what was or wasn't there before then.
After all, we know that Aquan, if it existed at all (my hypothesis may be utterly wrong!), did not leave attested descendants. Nowhere such a language is spoken now (unless Basque is one, but then Vennemann would be right and the family unrelated to IE), nor have inscriptions in such a language been found.
Of course, Etruscan could have been related to Palaeo-Hispanic languages (given the lack of knowledge, a distant relation wouldn't be readily apparent). Minoan or Eteocypriot (if they're not related to one another) might be related to Palaeo-Hispanic - we couldn't possibly know, until a lot more success has been had with those scripts. And of course a pre-IE language may have survived on Sardinia right the way into the historical era, but no attestations are known. Absence of evidence, however, is not evidence of absence.
Concurred. All that is possible.
I never claimed that all of Europe was ever as diverse as the Iberian Peninsula!
Then, again, what on earth are you "testing"? If you accept that the situation in Iberia tells us nothing at all about the situation outside of Iberia, how can Iberia be a "test case" of any theory?
This is the fourth time you ask this question in one single post ;)
But note that the Central Zone was not devoid of human inhabitants during the last ice age; in fact, it was rich in big game to hunt. Surely, more people moved in from the south when the ice age ended, the climate zones shifted and people followed their game northwards. So we have at least two linguistic layers - and at least two language families - in the Central Zone.
There have actually been several phases of replacement. As I understand it:

There were several phases before the glacial maximum. The earliest phases appear completely genetically unrelated to modern Europeans (more accurately: they are no more related to Europeans than to East Asians). Before the ice, Europe appears to have been genetically diverse.
When the ice came, the tundra region became more or less entirely depopulated, and the groups immediately preceding the ice do not appear to have had any genetic continuity with later groups.
As the ice receded, the big game hunters entered the region in much larger numbers. If there were any inhabitant before them, they became obsolete, leaving no apparent trace. The big game hunters were from Iberia (or southern France), and were related to one of the earlier groups, but not closely related to their immediate predecessors in the north.
The big game hunters brought about complete homogeneity in Europe, replacing the earlier diversity. At this point, remains everywhere from Spain to Hungary (although I don't know about the Balkans) became effectively genetically indistinguishable. The same culture was also rapidly spread over most of this area. It seems extremely plausible to me that this invasion spread a single language family, or at least no more than a couple, across the whole continent.
OK; I didn't know that. This would limit the linguistic diversity of Europe at that time.
Later, however, at the transition into the Mesolithic, another group arose. This group had some sort of connection to the Middle East, although it's too early to say whether they were immigrants, or emigrants, or just cousins. This new group rapidly spread throughout the whole continent; the earlier group left some traces, but were clearly genetically overwhelmed. Again, this new group brought with them their own distinctive culture. After this latest invasion and population replacement, Europe was again culturally and genetically homogenous. Again, it seems reasonable to suspect that it was also linguistically homogenous, or nearly so.
I see.
A few thousand years later, two big invasions occured from the middle east, associated with the introduction of farming. Although there were two invasions, the two groups seem genetically very close together. The earlier Mesolithic groups were marginalised; they left a substantial trace in most areas, but were clearly overwhelmed and replaced. [for one thing, Mesolithic Europeans were black (by modern standards), and Neolithic Europeans were not]. So again, after these invasions, Europe was genetically and (in two parts) culturally homogenous. It's not unreasonable to suspect that, perhaps outside of a few isolated Mesolithic holdout areas, it was also linguistically homogenous.
Indeed, I consider it likely that perhaps two language families spread across large swathes of Europe in the Neolithic, one north of the Alps, one in the Mediterranean, corresponding to the LBK and Cardial cultural lineages, respectively. The "LBK language family" could have been related to PIE and may have been my "Aquan" family.
Partway through the Neolithic, it seems like bad things happened. It was once thought that perhaps there was another invasion at this point. People don't think that anymore, I don't think - at least, if there were invasions, they were not massive population replacements. Maybe there were raiders? Probably more likely it's just that as the farmers spread out, they lost their cohesion and began to fall into conflict. From this point on, the genetic influence of the Mesolithic actually increases a little - some have suggested a "fightback", but it seems more likely this was just the gradual assimilation of the last holdouts, and that once they were assimilated they were actually sexually quite succesful.
Fair.
Then metalworking comes along. Some time later, there was a massive invasion from the Steppe by the Yamna (or a related group), who were partly distantly related to the mesolithics and partly not. Although their influence is less extreme in western and southern Europe (particularly in the Basque Country, where there is an unusally high level of mesolithic ancestry), the Yamna (or "kurgans") effectively overwhelmed the entire continent (apart from Sardinia). Again, Europe became fairly culturally and genetically homogenous (although not to the extent of earlier invasions). And Europe also became linguistically homogenous, so far as we can tell (and eventually it became almost entirely linguistically homogenous). This doesn't prove, of course, that the more complete earlier population turnovers didn't produce less linguistic domination, but it does seem, on the face of it, to strengthen the case somewhat...
This is just what most Indo-Europeanists think about the matter, and I am in full concurrence with them. Of course, the "Kurgan people" weren't much like Mongols, as is often popularly assumed. At least in the western part of their homeland, they were farmers, we have PIE words for the plough, and architectural terms which probably referred to more permanent structures than tents or yurts, also two words for pigs, an animal quite unsuitable for pastoral nomadism. And of course, the country they inhabited is famously suitable to agriculture, after all, Ukraine has been called "the granary of Russia" as long as it was part of the Muscovite empire and its "Communist" successor state.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

Addendum @Salmoneus (or whoever knows):

Where can I read up about the pre-Neolithic population movements you mentioned?
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

Duodecimal number system in "Old Europe"?

Francisco Villar claims, on page 155 of his 1996 book, Los Indo-Europeos y los orígines de Europa, that what he calls Vieja Europa 'Old Europe', i.e. the Neolithic peoples of the Balkan Peninsula, used a duodecimal number system, but doesn't give any sources. Does anybody have an idea where evidence for that can be found?
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

zompist
Boardlord
Boardlord
Posts: 3368
Joined: Thu Sep 12, 2002 8:26 pm
Location: In the den
Contact:

Re: European languages before Indo-European

Post by zompist »

WeepingElf wrote:Duodecimal number system in "Old Europe"?

Francisco Villar claims, on page 155 of his 1996 book, Los Indo-Europeos y los orígines de Europa, that what he calls Vieja Europa 'Old Europe', i.e. the Neolithic peoples of the Balkan Peninsula, used a duodecimal number system, but doesn't give any sources. Does anybody have an idea where evidence for that can be found?
That sounds really unlikely. I've looked at thousands of number systems; duodecimal is attested very very rarely.

I wonder if he's thinking of words like dozen / gross. But not every oddity in a language is a survival; it can be an innovation as well. ("Dozen" < Fr. douzaine < douze '12' < duodecim; "gross" < "grosse douzaine".) There's nothing amazing about having a special status for 12; it's a very useful number.

(Also, if you're tracing things back to the Neolithic, really extensive number systems should raise an eyebrow. Numbers are a technology, and hunter-gatherers usually don't have extendable number systems.)

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

zompist wrote:
WeepingElf wrote:Duodecimal number system in "Old Europe"?

Francisco Villar claims, on page 155 of his 1996 book, Los Indo-Europeos y los orígines de Europa, that what he calls Vieja Europa 'Old Europe', i.e. the Neolithic peoples of the Balkan Peninsula, used a duodecimal number system, but doesn't give any sources. Does anybody have an idea where evidence for that can be found?
That sounds really unlikely. I've looked at thousands of number systems; duodecimal is attested very very rarely.

I wonder if he's thinking of words like dozen / gross. But not every oddity in a language is a survival; it can be an innovation as well. ("Dozen" < Fr. douzaine < douze '12' < duodecim; "gross" < "grosse douzaine".) There's nothing amazing about having a special status for 12; it's a very useful number.
True. The number 12 is very useful and plays a major role in many cultures, but duodecimal numeral word systems are vanishingly rare. This is just one of the intriguing factoids Villar throws at the reader without giving any references. This utter lack of references renders the book almost entirely useless.

But in later works, Villar got worse still: in his 2012 book Indoeuropeos, Iberos, Vascos y sus Parientes, he puts geographical names in a blender, grinds them down to minute fragments like *il or *ur, draws isopleth maps of these resembling the maps of genetic markers geneticists draw, and constructs an elaborate linguistic stratigraphy involving up to six layers with the bottom layer being "archaeo-Indo-European". I think it is not just due to my limited proficiency in Spanish that I couldn't follow the argumentation.
(Also, if you're tracing things back to the Neolithic, really extensive number systems should raise an eyebrow. Numbers are a technology, and hunter-gatherers usually don't have extendable number systems.)
True. But the people concerned here weren't hunter-gatherers (otherwise, they wouldn't be Neolithic), but a quite sophisticated farming culture who, according to some scholars (though a minority) even may have had writing (see here for the symbol system in question). These people almost certainly did have a well-developed numeral system.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
Salmoneus
Sanno
Sanno
Posts: 3197
Joined: Thu Jan 15, 2004 5:00 pm
Location: One of the dark places of the world

Re: European languages before Indo-European

Post by Salmoneus »

I'll focus just on where I can be productive, rather than prolonging the argument...
Sure. But when half a dozen expansions of this kind happen in the same area one after the other, and in each case, remnants of the earlier landscape survive in residual zones such as mountain ranges or islands, the result can be quite some diversity.
That's a fair suggestion, I'll admit. I'm skeptical of it, however - after all, it requires that the survivors of each wave continue to survive each subsequent wave, while each wave also admits more survivors - up until the final wave, which left only one survivor. This seems rather implausible to me, but it is a possibility.
This is the first time I hear someone doubt that the Vascones of the Roman Era spoke anything like Basque! The Aquitanian language is only known from names occurring in Latin inscriptions of the Roman Era, but these names are, as I have been informed (R. L. Trask, The History of Basque, 1997), obviously Proto-Basque. Yet, I am no Vascologist and cannot evaluate whether Trask was right or wrong, but he was a major authority on the language. Is your knowledge of the relevant facts better than Trask's?
I don't doubt that Aquitanian was proto-Basque (or close to it). What I was questioning was whether the Vascones of that era spoke Aquitanian - after all, it's the Aquitanians who are said to have spoken Aquitanian, not the Vascones. And the Vascones were already established, with 'cities' of their own, at the time of the Aquitanians.

Our only link, so far as I'm aware, between the Basque and the Vascones is that later Romance speakers, centuries later, called the people we now call the Basque by a name derived from the old Latin word for the Vascones. This may indeed be because they knew them to be the descendents of the Vascones. But these links are very hard to rely on - after all, there's also a language called "Gascon", and that's not related to Basque!
[I think there are a few isolated funerary inscriptions and the like in something like Aquitanian in what is now Basque territory. But given that this was from the era of the Roman Empire when people moved around a lot, and that the Aquitanian territory was not far from Iberia, that doesn't seem too convincing as an attestation]

I may well be unaware of something obvious, however!

As I have clarified above, by "stock" I don't mean any kind of relationship however remote, but a family in a specific range of time depth, namely 4,000 to 8,000 years. Things like Nostratic or Dene-Caucasian, even if they were real, wouldn't be "stocks" because their time depths would fall far outside that range. It seems that we have been talking past each other here.[/quote]
I'm not sure how useful the concept "languages that last separated between 4 and 8 thousands years earlier" is. I don't see why it's useful. For one thing, it makes "number of stocks" a more or less arbitrary number based on finegrained splitting orders and timings that we can't possibly know about. You could have one stock sitting peacefully by itself in a region, then the clock ticks over the 4k mark and *bam* there are 32 different stocks there...
As I said above, "test case" may have been a misnomer on my side. All I meant was that by examining the relationships between the Paleo-Hispanic languages, one can get an idea of how many stocks there were, which could be helpful in guiding estimations of the linguistic diversity at least of the other two Mediterranean peninsulas.
But I don't see how that's possible - as you've said yourslef, it tells us nothing about the linguistic diversity of anywhere outside Iberia. Not that i'm dissuading you from demonstrating proto-Vasco-Iberian, mind you! Feel free to do so if you can! But now I think we're back to arguing, so I'll move on...
Fair. This is also what I think of the matter. And I am not alone. Francisco Villar said something similar in his book Los Indoeuropeos y los orígines de Europa - he considers the language of the Old European Hydronymy to be an early breakaway from PIE that was later clobbered by the historically known IE languages. That language would be a centum language (thus not Balto-Slavic) which merged PIE *o into *a (thus not Italo-Celtic) and did not shift its stops (thus not Germanic; the last point seems uncertain to me, though I am sure for other reasons I shall immediately come to that it wasn't Germanic). In my opinion, the dominant /a/-vocalism in the OEH is a problem here. Villar's idea raises the question: Why is everything in o-grade? To me, it looks like a language group that broke off before the rise of the PIE ablaut system, and preserved the pre-ablaut 3-vowel system. My ongoing work on a substratum dictionary seems to corroborates this hypothesis.
And, as I've said before, even assuming the existence of OEH, which seems unconvincing, I don't think that you can rely on precise vowel qualities (*e vs *a vs *o) being retained through thousands of years of borrowing and sound change, particularly in placenames. Indeed, what I'd expect is a gradual drift toward the most neutral quality, *a - which is what we see, so I don't think it's legitimate to use that to speculate on the morphological system of the putative ancestor...
An interesting idea - Basque coming from the east together with Indo-European. But why, then, don't we find other Vasconic languages elsewhere in Europe?
Well, why would we? We know that IE bulldozed every single language out of its way with the exception of Basque. That should remain equally true regardless of the linguistic affiliations of the prior languages.
[Now, if those Vasco-Caucasian crackpots could prove their case, that'd make it very interesting...]
Actually, Vasco-Caucasian may be correct - only the evidence adduced by people like Octaviano is shoddy. Basque and the two North Caucasian families may be the last remains of the same old layer of languages, alas, there seems to be no way knowing, and the hypothesis is based on superficial resemblances such as ergativity which prove nothing.
The idea of an old layer of languages seems highly implausible to me - as if everything were sitting around for so long waiting for the Indo-Europeans to come. Particularly given that there is no genetic affiliation between the two regions.

However, if Vasco-Caucasian WERE true, it would to me suggest that the Basque were indeed a fellow-traveller group of invaders of Europe; and the presence of such a caucasoid invading group could in turn explain the genetic presence of the caucasus in the Yamna, and the putatively pseudo-caucasoid character of PIE.

That said, I've no reason to believe that Vasco-Caucasoid IS true.

OK; I didn't know that. This would limit the linguistic diversity of Europe at that time.
Well, that's just may layman's understanding of the situation, I should say!
Indeed, I consider it likely that perhaps two language families spread across large swathes of Europe in the Neolithic, one north of the Alps, one in the Mediterranean, corresponding to the LBK and Cardial cultural lineages, respectively. The "LBK language family" could have been related to PIE and may have been my "Aquan" family.
As we've discussed previously, I think that this is impossible:
a) LBK originated in the middle east; the Yamnaya originated in the steppe. Neither appears genetically or culturally close to the other.
b) if the LBK did speak something IE-related, it would surely be far too far diverged from PIE (thousands of years!) to be so immediately recognisable.

If there is such a thing as Aquan/OEH, it would have to be spread by an unknown invasion from the Steppe. This is easiest to imagine occuring concurrently with, or immediately preceding, the main invasion. Alternatively, I guess you could imagine a non-demic invasion of the LBK area thousands of years before? To be honest, I think the more we find out about the past, the less likely non-demic elite invasions in prehistory look....


Regarding paleolithic and mesolithic Europe... I just read from time to time a few places online, particularly the eurogenes blog. Specifically, you might want to read Fu et al in Nature last year. I haven't, because I have neither the time nor the money. However, it's summarised on this blogpage
Blog: [url]http://vacuouswastrel.wordpress.com/[/url]

But the river tripped on her by and by, lapping
as though her heart was brook: Why, why, why! Weh, O weh
I'se so silly to be flowing but I no canna stay!

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

Salmoneus wrote:I'll focus just on where I can be productive, rather than prolonging the argument...
I see. You are a tough nut to crack - but so am I ;)
Sure. But when half a dozen expansions of this kind happen in the same area one after the other, and in each case, remnants of the earlier landscape survive in residual zones such as mountain ranges or islands, the result can be quite some diversity.
That's a fair suggestion, I'll admit. I'm skeptical of it, however - after all, it requires that the survivors of each wave continue to survive each subsequent wave, while each wave also admits more survivors - up until the final wave, which left only one survivor. This seems rather implausible to me, but it is a possibility.
True - we now have only one non-IE language left in western Europe, but 2,000 years ago, there were at least two more - Iberian and Etruscan - and perhaps a few more which we don't know of because they were never put to writing.
This is the first time I hear someone doubt that the Vascones of the Roman Era spoke anything like Basque! The Aquitanian language is only known from names occurring in Latin inscriptions of the Roman Era, but these names are, as I have been informed (R. L. Trask, The History of Basque, 1997), obviously Proto-Basque. Yet, I am no Vascologist and cannot evaluate whether Trask was right or wrong, but he was a major authority on the language. Is your knowledge of the relevant facts better than Trask's?
I don't doubt that Aquitanian was proto-Basque (or close to it). What I was questioning was whether the Vascones of that era spoke Aquitanian - after all, it's the Aquitanians who are said to have spoken Aquitanian, not the Vascones. And the Vascones were already established, with 'cities' of their own, at the time of the Aquitanians.

Our only link, so far as I'm aware, between the Basque and the Vascones is that later Romance speakers, centuries later, called the people we now call the Basque by a name derived from the old Latin word for the Vascones. This may indeed be because they knew them to be the descendents of the Vascones. But these links are very hard to rely on - after all, there's also a language called "Gascon", and that's not related to Basque!
[I think there are a few isolated funerary inscriptions and the like in something like Aquitanian in what is now Basque territory. But given that this was from the era of the Roman Empire when people moved around a lot, and that the Aquitanian territory was not far from Iberia, that doesn't seem too convincing as an attestation]

I may well be unaware of something obvious, however!
Surely, Gascon is unrelated to Basque, but we know that the area now known as Gascony was once Basque-speaking - the Aquitanian names are clearly Proto-Basque.
As I have clarified above, by "stock" I don't mean any kind of relationship however remote, but a family in a specific range of time depth, namely 4,000 to 8,000 years. Things like Nostratic or Dene-Caucasian, even if they were real, wouldn't be "stocks" because their time depths would fall far outside that range. It seems that we have been talking past each other here.
I'm not sure how useful the concept "languages that last separated between 4 and 8 thousands years earlier" is. I don't see why it's useful. For one thing, it makes "number of stocks" a more or less arbitrary number based on finegrained splitting orders and timings that we can't possibly know about. You could have one stock sitting peacefully by itself in a region, then the clock ticks over the 4k mark and *bam* there are 32 different stocks there...
I admit that this is a problematic concept, but how to measure linguistic diversity otherwise?
As I said above, "test case" may have been a misnomer on my side. All I meant was that by examining the relationships between the Paleo-Hispanic languages, one can get an idea of how many stocks there were, which could be helpful in guiding estimations of the linguistic diversity at least of the other two Mediterranean peninsulas.
But I don't see how that's possible - as you've said yourslef, it tells us nothing about the linguistic diversity of anywhere outside Iberia. Not that i'm dissuading you from demonstrating proto-Vasco-Iberian, mind you! Feel free to do so if you can! But now I think we're back to arguing, so I'll move on...
I am undecided about a Basque/Iberian connection. There are, as I said, seeming resemblances between the two languages, but not enough to establish a relationship.
Fair. This is also what I think of the matter. And I am not alone. Francisco Villar said something similar in his book Los Indoeuropeos y los orígines de Europa - he considers the language of the Old European Hydronymy to be an early breakaway from PIE that was later clobbered by the historically known IE languages. That language would be a centum language (thus not Balto-Slavic) which merged PIE *o into *a (thus not Italo-Celtic) and did not shift its stops (thus not Germanic; the last point seems uncertain to me, though I am sure for other reasons I shall immediately come to that it wasn't Germanic). In my opinion, the dominant /a/-vocalism in the OEH is a problem here. Villar's idea raises the question: Why is everything in o-grade? To me, it looks like a language group that broke off before the rise of the PIE ablaut system, and preserved the pre-ablaut 3-vowel system. My ongoing work on a substratum dictionary seems to corroborates this hypothesis.
And, as I've said before, even assuming the existence of OEH, which seems unconvincing, I don't think that you can rely on precise vowel qualities (*e vs *a vs *o) being retained through thousands of years of borrowing and sound change, particularly in placenames. Indeed, what I'd expect is a gradual drift toward the most neutral quality, *a - which is what we see, so I don't think it's legitimate to use that to speculate on the morphological system of the putative ancestor...
An interesting idea - Basque coming from the east together with Indo-European. But why, then, don't we find other Vasconic languages elsewhere in Europe?
Well, why would we? We know that IE bulldozed every single language out of its way with the exception of Basque. That should remain equally true regardless of the linguistic affiliations of the prior languages.
I consider it unlikely that earlier waves of language family expansion were just as devastating as the the combination of IE, Celtic and Latin expansions which shaped the modern linguistic landscape of western Europe, as they probably did not have the military prowess of the Proto-Indo-Europeans or the Celts, let alone the Roman Empire. The landscape immediately before Indo-Europeanization may have been dominated by one or two large families that spread together with agriculture, but there probably were a few residues of the Mesolithic linguistic landscape, I think.
[Now, if those Vasco-Caucasian crackpots could prove their case, that'd make it very interesting...]
Actually, Vasco-Caucasian may be correct - only the evidence adduced by people like Octaviano is shoddy. Basque and the two North Caucasian families may be the last remains of the same old layer of languages, alas, there seems to be no way knowing, and the hypothesis is based on superficial resemblances such as ergativity which prove nothing.
The idea of an old layer of languages seems highly implausible to me - as if everything were sitting around for so long waiting for the Indo-Europeans to come. Particularly given that there is no genetic affiliation between the two regions.
It's not that I consider the reality of Vasco-Caucasian likely. The "evidence" I have seen is laughable.
However, if Vasco-Caucasian WERE true, it would to me suggest that the Basque were indeed a fellow-traveller group of invaders of Europe; and the presence of such a caucasoid invading group could in turn explain the genetic presence of the caucasus in the Yamna, and the putatively pseudo-caucasoid character of PIE.

That said, I've no reason to believe that Vasco-Caucasoid IS true.
Nor do I.

OK; I didn't know that. This would limit the linguistic diversity of Europe at that time.
Well, that's just may layman's understanding of the situation, I should say!
Indeed, I consider it likely that perhaps two language families spread across large swathes of Europe in the Neolithic, one north of the Alps, one in the Mediterranean, corresponding to the LBK and Cardial cultural lineages, respectively. The "LBK language family" could have been related to PIE and may have been my "Aquan" family.
As we've discussed previously, I think that this is impossible:
a) LBK originated in the middle east; the Yamnaya originated in the steppe. Neither appears genetically or culturally close to the other.
b) if the LBK did speak something IE-related, it would surely be far too far diverged from PIE (thousands of years!) to be so immediately recognisable.

If there is such a thing as Aquan/OEH, it would have to be spread by an unknown invasion from the Steppe. This is easiest to imagine occuring concurrently with, or immediately preceding, the main invasion. Alternatively, I guess you could imagine a non-demic invasion of the LBK area thousands of years before? To be honest, I think the more we find out about the past, the less likely non-demic elite invasions in prehistory look....
I have to admit that I am not all too knowledgeable about European paleogenetics. I am not convinced myself that there is a connection between LBK and IE. Yet, it seems to me that a language family related to IE was once widespread in western and central Europe, manifesting in the Old European Hydronymy and in substratum loanword in western IE languages. But maybe I have been misled by delusionary evidence, and I have just been hunting linguistic ley lines.
Regarding paleolithic and mesolithic Europe... I just read from time to time a few places online, particularly the eurogenes blog. Specifically, you might want to read Fu et al in Nature last year. I haven't, because I have neither the time nor the money. However, it's summarised on this blogpage
Thank you.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

Regarding the genetic discrepancy you mention between LBK and Yamna (source?), this of course does not prove that their languages were unrelated. Languages can spread from one genetic population to another. The Sami, I have heard, are genetically not much like Finns, yet they speak a related language. The whole Uralic and Turkic families straddle the boundary zone between "European" and "East Asian" human "races" (I know that "race" is a problematic concept avoided by modern geneticists). Compare Finns to Samoyeds, or Turks (from Turkey) to Yakuts.

It is of course the question which way the language spread. It seems as if the whole shebang is more distantly related to Uralic, but that is not certain. That seems to indicate that the language came from the east. There may have been a component in the ancestors of the LBK which brought in a sister language of PIE but left only a minor mark on the gene pool. I need to do more research on these matters.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
mèþru
Smeric
Smeric
Posts: 1984
Joined: Thu Oct 29, 2015 6:44 am
Location: suburbs of Mrin
Contact:

Re: European languages before Indo-European

Post by mèþru »

Turks are genetically mostly the same as the pre-Turkic peples of Anatolia and Greece. Also, there doesn't seem to be enough evidence about the history of non-Indo-European languages to either confirm or deny most of what WeepingElf says. However, Aquitanian seems to be pretty widely accepted being related to Basque, although it is unknown whether Aquitanian is a sister language to Proto-Basque or is synonymous with Proto-Basque.
WeepingElf wrote:I admit that this is a problematic concept, but how to measure linguistic diversity otherwise?
Accept that there is not enough evidence and move to a new topic. New evidence might be found, but until then there is no point.
ìtsanso, God In The Mountain, may our names inspire the deepest feelings of fear in urkos and all his ilk, for we have saved another man from his lies! I welcome back to the feast hall kal, who will never gamble again! May the eleven gods bless him!
kårroť

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: European languages before Indo-European

Post by WeepingElf »

WeepingElf wrote:Regarding the genetic discrepancy you mention between LBK and Yamna (source?), this of course does not prove that their languages were unrelated. Languages can spread from one genetic population to another. The Sami, I have heard, are genetically not much like Finns, yet they speak a related language. The whole Uralic and Turkic families straddle the boundary zone between "European" and "East Asian" human "races" (I know that "race" is a problematic concept avoided by modern geneticists). Compare Finns to Samoyeds, or Turks (from Turkey) to Yakuts.

It is of course the question which way the language spread. It seems as if the whole shebang is more distantly related to Uralic, but that is not certain. That seems to indicate that the language came from the east. There may have been a component in the ancestors of the LBK which brought in a sister language of PIE but left only a minor mark on the gene pool. I need to do more research on these matters.
I have meanwhile read up a bit on Neolithic European human genetics. It seems that LBK and Yamnaya were genetically very different from each other, with modern Europeans more closely resembling the latter than the former. I also no longer maintain the notion that the hypothetical Proto-Aquan language was that of the LBK people. Rather, it may have been a part of the first "Kurgan" advance ca. 4500 BC, which would still make it an early offshot of IE, perhaps Para-Anatolian (i.e., more closely related to Anatolian than to the rest of the family), perhaps an earlier branch than Anatolian. Maybe the absence of PIE ablaut is secondary due to the following sound changes:

1. PIE *a, *e and *o merge as *a (as in Indo-Iranian).
2. PIE zero grades develop a-vocalism.
3. PIE vowel length is lost.
4. PIE diphthongs are monophthongized as *i and *u.

Voilà, ablaut is gone. With just four simple sound changes that do not look out of the ordinary at all.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

Post Reply